Major depressive disorder (MOD) is common, recurrent, and disabling. Despite a wide range of antidepressant treatments available, MOD remains difficult to treat. Therefore, a better understanding of treatment mechanisms and identification of reliable predictors of treatment response would constitute major progress in the battle against MOD. Recently, there has been considerable interest in studying the efficacy of alternative medications in the treatment of MOD. In this regard, S-adenosyl-L-methionine (SAMe) has been widely investigated as a natural treatment of depression. Although preclinical studies have shown that SAMe influences neurotransmitter metabolism (including dopamine, a neurotransmitter implicated in reward brain mechanisms), membrane fluidity, and receptor activity, the precise mechanisms of its antidepressant actions are unknown in humans. As an initial step toward a mechanistic understanding of these processes, the investigators propose to compare the effects of SAMe to a standard, first-line treatment for depression (the selective serotonin reuptake inhibitor, escitalopram) on neural activity underlying reward processing in MOD. By using functional magnetic resonance imaging (fMRI) in conjunction with a reward task (a monetarily-reinforced button-press task) as well as a laboratory-based behavioral measure of hedonic capacity, this study will investigate: (1) the effects of SAMe and escitalopram on neural substrates underlying processing of reward- and punishment-related cues; (2) pre-treatment behavioral and neural predictors of treatment response to SAMe and escitalopram; and (3) putative differences in the effects of SAMe and escitalopram - due to their distinct pharmacological profiles - on neural and behavioral markers of hedonic capacity. To this end, 45 individuals with MOD (and 15 healthy control subjects) will be investigated (before and after a 12-week, double-blind, placebo-controlled administration of SAMe vs. escitalopram. Based on preclinical evidence suggesting that SAMe potentiates transmission in the mesolimbic dopaminergic reward system, we hypothesize that, compared to placebo, SAMe (and escitalopram) treatment will: (1) normalize (i.e., increase) brain activation in regions subserving reward processing (ventral striatum, orbitofrontal cortex, anterior cingulate cortex); (2) normalize (i.e., decrease) brain activation in regions subserving punishment processing (e.g., amygdala, insula, hippocampus, right dorsolateral prefrontal cortex); and (3) increase hedonic capacity, as assessed by a signal-detection task. Because anhedonia (lack of reactivity to pleasurable stimuli) has been considered a trait marker related to vulnerability to depression and a predictor of relapse, the use of a novel experimental approach capitalizing on state-of-the-art neuroimaging techniques and laboratory-based measures of hedonic capacity promises to shed important light on: (1) mechanisms underlying the antidepressant efficacy of a natural treatment for depression; and (2) objective predictors of treatment response. A better understanding of the mechanisms of actions of SAMe should, in turn, provide a foundation for future clinical trials.